Inorg. Chern. 1986, 25, 3551-3552
3551
are available as with the ligand ir to metal eg* states of 7.0 C ~ ( p h e n ) and ~ ~ +C ~ ( b p y ) , ~ +Both . C O ( N H & ~ +and C ~ ( e n ) ~ ~ + possess two LF bands. Photoexcitation of either L F band leads to a high degree of DNA cleavage. Clearly neither Co(NH3):+ nor C ~ ( e n ) ~bind ~ + to D N A via intercalation. Indeed these complexes are known to maintain intimate hydrogen-bonding interactions with the helix.23 Again, coupling inefficient photoreactions to DNA binding of all sorts may yield useful photo5.0 cleavers. These photoactivated cleavage reactions, in particular those with C O ( N H ~ ) ~might ~ + , be helpful in mapping and phoX i o 4 M tofootprinting experiments. These results indicate that light-induced cleavage of DNA by a range of d6 metal complexes can occur from CT, LF, and IL states of simple inorganic complexes, albeit via different mechanisms. Cleavage occurs with complexes having positive charges 3.0 of either 2+ or 3+. The complexes cleave D N A in spite of 0 200 400 possessing different binding modes to the DNA (intercalative, time (sec) hydrogen bonded, and electrostatic). It may therefore be possible to couple the binding mode of interest with the photoactivated Figure 1. Time dependence of [I-] during and after 546-nm photolysis cleavage process of the metal complexes to give a variety of of [Cr(NH3),Il2+in acidic aqueous perchlorate medium at 7 OC: points, experimental data; solid line, theoretical fit (see text). Two data sets are site-specific DNA cleavers. Photoactivated cleavage of DNA with shown corresponding to the best and worst fits found, one displaced 1.2 simple coordination complexes may be considered a general X M upward for clarity. Photolysis begins at zero and ends at the phenomenon. Moreover, since these reagents are activated by arrow. near-UV and visible light, they may provide useful probes for mapping experiments in vivo. the stereochemical results required some extensions.'s As has Acknowledgment. We gratefully acknowledge the fiiiancial been pointed out,I6 the results for the bromo pentaammine comsupport of the Army Office of Research (to N.J.T.), the National pound are somewhat equivocal since the calculated excited-state Science Foundation (Grant CHE-83-51017; J.K.B. is a Presibond strengths for the Cr-Br and Cr-N bonds are nearly equal, dential Young Investigator), and the National Institutes of Health but the molecule efficiently loses mainly ammonia at all wave(Grants GM33309 to J.K.B. and CA07957 to K.C.W.). We also lengths. thank Howard Oster for expert experimental assistance. Surprisingly the iodo compound has never been studied photochemically, although it has been reported and its thermal Department of Chemistry Mindy B. Fleisher aquation studied p~larographically.'~Probably its lability has Columbia University Kenneth C. Waterman deterred photochemical study, the half-life for iodide aquation New York. New York 10027 Nicholas J. Turro* in 25 OC aqueous solution being only 17 min. Jacqueline K. Barton* It is important, however, to fill this gap in our knowledge and to take advantage of the potential offered by this molecule to Received May 30. I986 provide a useful test of the VC theory. Using as ligand field parameters, for ammonia, E,, = 0.7183 pm-' and, for iodide, E , = 0.4583 pm-' and E, = 0.065 wm-', the VC theory predicts, for the 4E, lowest excited quartet state, bond energies of 0.72, 0.82, and 1.21 pm-' for the Cr-I, Cr-NH3(ax), and Cr-NH,(eq) bonds, Photoaquation of the Iodopentaamminechromium(II1) Ion respectively, unambiguously implying dominant photochemical in Acidic Aqueous Solution loss of iodide ion in contrast to the Adamson rule prediction of ammonia loss. We have therefore chosen to study this molecule, Sir: recognizing that the problems deriving from the thermal lability The acidopentaamminechromium(II1) complexesi-8 [Crof the complex would limit the precision of any results and (NH3)5X]2+where X = Cl-, Br-, and F, together with some of probably obscure beyond recovery the finer details of the wavetheir analogues containing ethylenediamine in place of four amlength and temperature dependence of the yields as well as the manias,*-" have figured importantly in studies of ligand labilization photostereochemistr y. and stereochemistry of the photosubstitution reactions of d3 [Cr(NH3)51]12was prepared as previously describedi7 and complexes. characterized by its visible spectrum, which showed bands at 532 For the large majority of such complexes the photochemistry is well accounted for by the angular overlap ligand field m ~ d e l ' ~ . ' ~ nm (38.1 L mol-' cm-I) and 472 (36.4), compared with the literature" values 530 (40) and 472 (40). Chromium analysis by of Vanquickenborne and Ceulemans (hereafter VC). For the oxidation with boiling alkaline hydrogen peroxide and spectrofluoro compound particularly the theory gave a better description photometric determination of chromate gave 10.3 & 0.25% (five of the photosubstitution modes than Adamson's ruleI4 although runs), consistent with theory, 10.04%. Irradiation at 8 'C and 546 nm, into the first ligand field quartet Moggi, L.; Bolletta, F.; Balzani, V. Ric. Sci. 1966, 36, 1228. band, of 5 X M solutions in 1.319 X M HC10,/0.04 Wasgestian, H. F.; Schlaefer, H. L. Z . Phys. Chem. (Munich) 1968, M KCIOl aqueous medium led to a rise in pH, demonstrating 57, 282. photoaquation of ammonia. Reineckate actinometry together with Wasgestian, H. F.; Schlaefer, H. L. Z . Phys. Chem. (Munich) 1968, absorbance measurements and standard additions of H + deter62, 127. Manfrin, M. F.; Moggi, L.; Balzani, V. Inorg. Chem. 1971, 10, 207. mined a quantum yield of 0.28 f 0.04 (1 3 runs) for this process. Wong, C. F. C.; Kirk, A. D. Can. J . Chem. 1974, 52, 3384. The thermal lability of the compound caused unusually large Wong, C. F. C.; Kirk, A. D. Can. J . Chem. 1975, 53, 419. scatter in these data and, more importantly, precluded reliable Riccieri, P.; Schlaefer, H. L. Inorg. Chem. 1970, 9, 727.
1-
Wong, C. F. C.; Kirk, A. D. Inorg. Chem. 1977, 16, 3148. Wright, R. E.; Adamson, A. W. Inorg. Chem. 1977, 16, 3360. Wong, C. F. C.; Kirk, A. D. Inorg. Chem. 1976, 15, 1519. Kirk, A . D. Inorg. Chem. 1979, 18, 2326. Vanquickenborne, L. G.; Ceulemans, A. J . Am. Chem. SOC.1977, 99,
2208.
Vanquickenborne, L. G.; Ceulemans, A. J . Am. Chem. Sor. 1978, 100, 475.
0020-1669/86/1325-3551$01.50/0
(14) Adamson, A. W. J . Phys. Chem. 1967, 7 1 , 798. (15) Vanquickenborne, L. G.; Ceulemans, A. Inorg. Chem. 1979, 18, 897, 3475. (16) Endicott, J. F. J . Chem. Educ. 1983, 60, 824. (17) Levine, M. A,; Jones, T. P.; Harris, W. E.; Wallace, W. J. J . Am. Chem. SOC.1961, 83, 2453.
0 1986 American Chemical Society
Inorg. Chem. 1986,25, 3552-3554
3552
correction for effects owing to the concomitant thermal formation of [Cr(NH3)5H20]3+.Here the important effect is a reduction of the fraction of light absorbed by the iodo complex, and hence the measured quantum yield should be considered a reliable lower limit. The true value may well be closer to the value of 0.35 common to the chloro2 and bromo’ analogues. Loss of iodide was considerably more difficult to follow for several reasons. First, we found it impossible to obtain either the perchlorate or the nitrate salts by reprecipitation techniques owing to the low solubility of the iodide salt and rapid decomposition of the complex ion. Photochemical iodide loss had therefore to be investigated in solutions from which the iodide ion had been removed by anion exchange through a 3 cm long column (perchlorate form of IRA 400 resin, 100 mesh) near 8 OC. This resulted in a solution of ill-defined composition containing the iodo pentaammine and aquo pentaammine ions. Were time now committed to a spectrophotometric determination of the composition, aquation was subsequently so advanced as to preclude useful photochemical measurements. It was best therefore to directly follow the time course of iodide formation at 7 “ C before, during, and after irradiation by using a concentration cell with Ag/AgI electrodes. The resulting curves (Figure 1) show three phases: (1) thermal aquation of iodide prior to photolysis; the measured rate constant (7 “C) for our conditions was 1.23 X s-I, in s-’ M calculatedI8 good agreement with the value of 1.0 X from k,,, and E,,, = 21.4 kcal mol-,; (2) release of additional iodide during the irradiation period, leading to upward curvature in the iodide/time plot; (3) thermal release of iodide at an increased but declining rate after photolysis. These observations show that any photochemical iodide loss is obscured not only by the thermal loss of iodide from the iodo complex but also, and worse, by rapid thermal iodide aquation from the primary photoproduct of the ammonia loss mode, almost certainly c i ~ - [ C r ( N H ~ ) ~ H , 0 ( 1 ) ]This ~ + . phenomenon has been observed previously for the chloro c o m p o ~ n d . ~ .To ’ ~ determine any photochemical quantum yield for iodide therefore, the starting concentrations, the absorbances at 546 nm of all species, the rate constants for the significant thermal reactions, and the light intensity and ammonia quantum yield are required. Five of these were unknowns, namely the exact initial concentrations of the iodo and aquo pentaammines, the iodide aquation rate constant and molar absorptivity at 546 nm of the major photoproduct, and the quantum yield for direct photoaquation of iodide. The approach taken was to use the SIMPLEX method with a least-squares criterion for goodness of fitm to find the parameters in the differential equations for the main thermal and photochemical processes. These equations are d[I-]/dt = kl[CI]
+ k,[CWI] + 421a
d[CW]/dZ = kl[CI] +421a
were superior to those where 42 was 0.1, consistent with this. Figure 1 shows the experimental points with fits (solid lines) for the best and worst data sets that we examined in detail. We conclude the following: (1) For all data sets fits can be found having relative standard deviations of 1 4 % and giving consistent and physically reasonable parameters. The deviations from experiment are clearly not random but can be partially explained by hysteresis in the iodide electrode response. (2) The calculated initial concentrations correspond well to a slightly diluted and partially aquated (10-30%) version of the original solution placed on the ion-exchange column. (3) The rate constant for iodide loss from the ammonia phos-l (four runs), about 30 times toaquation product is (3 f 1) X that for the iodo pentaammine. This factor compares with a factor of about 100 found for the analogous chloro pentakis(methy1amino) c o m p o ~ n d . ’ ~ (4) The “best fit” quantum yield for iodide photolysis is close to zero. Our results show that the iodo pentaammine, like the chloro and bromo pentaammines, loses ammonia efficiently on irradiation into the ligand field bands with at most a minor yield of iodide. Its photochemistry is therefore also antithermal, inconsistent with the predictions of the VC theory and consistent with Adamson’s simpler rule. In terms of fundamental behavior, however, we do not make much of this last point. It seems that the excited-state bond strength, assuming the VC calculated values may be relied upon, cannot be the only important factor operating. Many years ago, Manfrin et al. pointed out4 that much of Cr(II1) photochemistry was consistent with another simple rule, preferential aquation of an uncharged ligand. An enhanced escape rate of the neutral ammonia molecule probably accounts for the deviation of this molecule (and the bromo pentaammine) from the VC theoretical predictions. Acknowledgment. The University of Victoria and the Natural Sciences and Engineering Research Council of Canada generously provided financial support. Department of Chemistry University of Victoria Victoria, BC, Canada V8W 2Y2
A. D. Kirk* S. Arunachalam D. Kneeland
Received February 20, 1986
Synthesis and Characterization of Substitution-Inert Cobalt(111) Complexes Containing an N-Glycoside Derived from Ethylenediamine and a Monosaccharide (D-Mannose, L-Rhamnose, or D-Ribose)
Sir: where [CI] = [[Cr(NH3)51]2+],[CWI] = [[Cr(NH3),H2O(1)I2+], We report the first successful synthesis, isolation, and char[CW] = [[Cr(NH,),H,0]3+], and I, = I&(l - IO-*’) with Abs acterization of substitution-inert cobalt(II1) complexes containing the total absorbance of all species at 546 nm,fcr the fraction of an N-glycoside derived from ethylenediamine (en) and D-mannose absorbed light absorbed by the iodo compound, k , and k2 the (D-Man), L-rhamnose (L-Rha), and D-ribose (D-Rib), the results iodide aquation rate constants for CI and CWI, respectively, and of which suggest a novel coordination behavior of sugars. $ I and 42 the ammonia and iodide quantum yields from CI. The interaction of carbohydrates with transition metals is of Fortunately not all five unknown parameters had to be varied current interest in coordination chemistry and in bioinorganic in an unconstrained way. Initial simulations showed that the results were fairly insensitive to the exact value of c ~ ~ ~ ( C W I ) . chemistry. Despite a wide interest in complexes with sugars, very little is known about the structural details of the complexes.’ Therefore, on the basis of comparison with the spectra of [CrMetal complexes are divided into two classes such as substitution C I )unknown ]*+, was (NH3)sC1]2+and C ~ S - [ C ~ ( N H ~ ) ~ H ~ O (this inert and substitution labile on the basis of their chemical propset at 40 L mol-, cm-I. Simulations with the remaining four erties. Therefore, it is desirable to examine these two types of unknowns unconstrained led to negative quantum yields of iodide complexes in order to elucidate the interaction of sugars with for all four data sets studied. Best fits with d 2 constrained at 0.0 metals. Recently, we have extensively investigated the synthesis (18) Parris, M.; Wallace, W. J. Inorg. Chem. 1964, 3, 133. (19) Wong, C. F. C.; Kirk, A. D. Can. J. Chem. 1974, 52, 3384. (20) Shavers, C. L.; Parsons, M. L.; Deming, S. N. J. Chem. Educ. 1979, 56, 307.
0020-1669/86/1325-3552$01 S O / O
(1) (a) Angyal, S. J. Chem. SOC.Reu. 1981, 9, 415-428. (b) Hanessian, S.; Patil, G. Tetrahedron Lett. 1978, 1031-1034. (c) Hanessian, S.; Patil, G. Ibid. 1978, 1035-1038.
0 1986 American Chemical Society
